基于BP神经网络的涡轮增压机组压气机特性计算

Characteristic Calculating of Compressor in Turbo-charger Set Based on BP Neural Network

在涡轮增压机组变工况热力计算中,因为涉及到需要依据输入数据反复计算压气机特性参数的问题,使得整个计算过程复杂繁琐、效率低下。应用BP神经网络求出压气机特性曲线的函数关系,并将其引入到Excel中,可以实现机组的变工况热力计算。实例表明,该计算表格的查值效率较高、通用性较好,可推广至其它涉及到压气机特性计算的地方。

In the change-condition calculation of turbo-charger set,the problem which refers to repeat characteristic calculating of compressor according to input data have made the whole calculating process complicated and low efficiency.Applying BP neural network to working out the functional relation in the characteristic map of compressor,and then introduces it into Excel for the change-condition characteristic calculating of turbo-charger set.A given example showed that the efficiency and universality of the Excel for characteristic calculating of compressor is remarkable,and that can be expanded to the other place where the characteristic calculating of compressor involved.